Anthropic Accuses Alibaba Of Large-Scale AI Theft, Claims Chinese Giant Used Almost 25,000 Fraudulent Accounts To Generate 28.8 Million Exchanges With Claude
· Free Press Journal
Anthropic has formally accused Chinese technology conglomerate Alibaba of orchestrating what it calls the largest known distillation attack against its Claude AI models to date, escalating a dispute that AI labs have so far waged largely against smaller Chinese startups rather than a corporate giant of Alibaba's scale.
What is Anthropic accusing Alibaba of?
Visit saltysenoritaaz.com for more information.
In a letter addressed to Senate Banking Committee Chairman Tim Scott and ranking member Elizabeth Warren, Anthropic's head of policy, Sarah Heck, alleged that Alibaba used almost 25,000 fraudulent accounts to generate more than 28.8 million exchanges with Claude, targeting some of the model's most valuable capabilities such as agentic reasoning, software engineering and long-horizon tasks. The activity is alleged to have run between April 22 and June 5, and was reportedly conducted by operators affiliated with Alibaba and its Qwen AI lab.
Anthropic Files For IPO Confidentially, Racing Rival OpenAI & SpaceX To Wall Street This FallThe practice Anthropic describes, distillation, involves querying a rival company's model through public APIs at industrial scale and using its outputs to train a competing, typically cheaper, model. Anthropic specifically claimed the campaign was designed to help Alibaba's Qwen model access advanced capabilities from Anthropic's frontier 'Mythos Preview' model. The company did not mince words about the stakes, warning that such efforts "turn hundreds of billions of dollars in American investment and R&D into a massive subsidy for our geopolitical competitors."
Anthropic also noted this wasn't an isolated incident. It pointed out that the campaign follows the same pattern it disclosed in February involving DeepSeek, Moonshot and MiniMax, which combined had generated over 16 million exchanges with Claude through roughly 24,000 fraudulent accounts. The company further claimed Alibaba pressed ahead even after a White House memo in April had warned foreign entities about industrial-scale efforts to harvest US AI systems.
Anthropic Flies Staff To Washington To Patch Up White House DisputeBeyond the accusation itself, Anthropic used the letter to push for policy action, urging Congress to pursue stricter export restrictions on chips and to pass legislation penalising AI developers caught running distillation operations. Bloomberg was first to report on the existence and contents of the letter.
The broader context
The accusation lands at a particularly fraught moment for Alibaba's standing in Washington. The Pentagon added Alibaba to its Section 1260H list of Chinese military companies on June 8, and Alibaba filed a federal lawsuit in the Northern District of California on June 23 challenging that designation. Alibaba's American depositary receipts fell roughly 2.7 percent to close at $99.80 on the day Anthropic's letter became public, though separate reporting suggested the market's reaction to the distillation claims specifically was relatively muted, with shares dropping 4.4 percent in Hong Kong trading the following day, underperforming the Hang Seng Tech Index's 1.6 percent decline. One analyst told reporters that given this isn't the first distillation allegation targeting Chinese firms, the reputational damage was expected to be small.